Preliminary study of food habits in the Japanese clawed salamander larvae (Onychodactylus japonicus) in a mountain brook of the Kiso River system
Teruhiko Takahara1,2*, Motomi Genkai-Kato1,3, Hitoshi Miyasaka1 and Yukihiro Kohmatsu1,2
1 Kiso Biological Station, Kyoto University (6388 Fukushima, Kiso, Nagano 397-0001)
2 Research Institute for Humanity and Nature (457-4 Motoyama, Kamigamo, Kita-ku, Kyoto 603-8047)
3 Graduate School of Kuroshio Science, Kochi University (2-5-1 Akebono-cho, Kochi 780-8520)
Abstract
To evaluate food habits of the Japanese clawed salamander larvae (Onychodactylus japonicus), we exam- ined stomach contents of 22 individuals collected from a natural mountain brook in a tributary of the Kuro- kawa River in Kiso Fukushima, Nagano Prefecture, central Japan. Their diet composition did not differ between fast and slow current conditions. The diet reflected the natural benthos communities of the brook, in which mayfly nymphs and caddisfly larvae accounted for 70–88%. The salamander larvae selectively foraged on caddisfly and fly larvae in the two current conditions. In the slow current condition, their prey preference was detected in the prey group including terrestrial invertebrates. Our results suggest that the prey items of O. japonicus larvae hinge primarily on the natural benthos communities.
Key words: benthic animals, habitat traits, Manly’s Index, prey preference, terrestrial invertebrates
Introduction
Diverse food items found in animal stomachs rep- resent the ability to utilize a wide variety of prey taxa. In fact, knowledge of the feeding habits of aquatic animals, such as aquatic insects, fish and amphibians, can provide useful system-specific information on variation in feeding behavior, food availability, and prey species diversity (e.g., Nakano et al., 1999; Hirai and Matsui, 2000;
Matsui et al., 2003; Ohba and Nakasuji, 2006).
Lotic habitats are characterized by heterogeneous physical conditions, even within a local area (Nowell and Jumars, 1984). The action of flowing water is a dominant physical feature of river systems, which mediates physi- ological processes such as temperature and oxygen avail- ability and influences the distribution and abundance of benthic animals (e.g., Genkai-Kato et al., 2000, 2005).
Water flow affects interactions among organisms such as grazer–periphyton (Poff and Ward, 1995; Kuhara et al., 2000) and predator–prey interactions (Hansen et al., 1991).
The Japanese clawed salamander, Onychodactylus japonicus (Houttuyn) (Fig. 1), is distributed in Honshu and Shikoku Islands of Japan and is found in montane
regions ranging from a few hundred meters to over 2000 m above sea level (Goris and Maeda, 2004). Salamander larvae inhabit lotic environments, such as brooks and streams, for two to three years until metamorphosed (Ogihara and Nakamura, 1974; Hayase and Yamane, 1982). The presence of salamander larvae is a symbol of clear water in aquatic systems. Thus, salamanders are of importance for conservation. In recent years, however, they have decreased in abundance as shown in other amphibian species (Stuart et al., 2004), and are listed as a near threatened species or vulnerable species in some prefectures of Japan (e.g., Sugiyama, 2004).
The dietary items and habitat traits can be closely
Fig. 1 J a p a n e s e c l a w e d s a l a m a n d e r l a r v a e (Onychodactylus japonicus)
Received 13 May 2010；accepted 24 November 2010.
* Corresponding author: e-mail [email protected]
related with the survival of salamander. Stomach contents are one of the most relevant features of the life history of amphibians (Ihara, 1998; Blackburn and Moreau, 2006;
Wheeler et al., 2007), although direct observations of foraging behavior in the field are difficult. There is no quantitative or qualitative report on the food items of O.
japonicus larvae in natural environments. Here, we exam- ined habitat traits and prey compositions in stomachs of O.
japonicus larvae in a mountain brook of Japan.
1. Methods
A total of 22 salamander larvae were collected on 30 July 2001 from Akashio-zawa Brook (35˚52´N, 137˚40´E) in Kiso Fukushima, Nagano Prefecture, cen- tral Japan (Fig. 2). The brook is a tributary of Kuro-kawa Stream, flowing into the Kiso River system (Fig. 3). The Kiso River system flows finally into the Pacific Ocean through Ise Bay. Sampling of salamander larvae was conducted with a quadrat net (10 × 8 cm, <1-mm mesh) by randomly collecting a stone (loose cobble) from the streambed of the Akashio-zawa Brook. We measured the snout–vent length of collected individuals as an indi-
cator of body size to the nearest 0.01 mm using a digital caliper (Code No. 500-151, Mitutoyo, Kanagawa, Japan).
To assess habitat traits with regard to physical condi- tions, we measured the current velocity and water column depth at each site where a larva was collected. We also measured the surface area of the stone when a larva was collected. Current velocity was measured at 60% depth to estimate the mean current velocity (Hynes, 1970), using a portable current meter (Model CR-7WP, Cosmo-Riken,
Fig. 2 Study site (open circle) of Akashio-zawa Brook in Kiso Fukushima, Nagano prefecture, and the Kiso River system in central Japan.
Fig. 3 Akashio-zawa Brook (a, b), and Kuro-kawa Stream into which the brook flows (c, d).
Aichi Pref.
Ise bay Gifu Pref.
Nagano Pref.
Lake Suwa
Shizuoka Pref.
Kiso River
25km Kuro-kawa River
Study site Ｎ
(a)
(c)
(b)
(d)
Kashihara, Japan). Water temperature in the sampling brook was measured on the sampling day. The collected salamanders were killed immediately by immersion in 5% formalin to preserve the stomach contents. They were then transported to the laboratory of the Kiso Biological Station, Kyoto University. The stomach contents were examined by abdominal dissection under a binocular microscope.
To assess the benthos communities as potential prey items for the salamander larvae in the brook, we col- lected benthic animals with a quadrat net (15 × 15 cm,
<1-mm mesh) at fast and slow current sites (3 replicates at each site). Benthic animals were categorized into mayfly nymphs, caddisfly larvae, stonefly nymphs, fly larvae, and others (e.g., dragonfly nymphs and terrestrial invertebrates). The proportional contribution of each prey category to the total stomach contents or the total benthos communities in the brook was determined for each cur- rent condition.
The physical conditions of the microhabitats were compared between two current conditions with t-tests.
Animal prey compositions were compared between two current conditions in the field survey and stomach con- tents using multivariate analysis of variance (MANOVA).
Proportional data were arcsine-square-root transformed.
The Manly’s Index (Chesson, 1978) was used as a mea- sure of prey selectivity:
m i
n r
n
mr
j j j
i
i i , 1,2,...,
/ /
1
=
=
∑
=α
where ri and ni are the proportions in biomass of prey category i occurring in stomach contents and field survey, respectively. The total number of prey item categories, m, is 5 in this study. This index, which removes bias due to differing availability of prey categories, lies between 0 (never eaten) and 1 (100% eaten). A preference is indi- cated by values of αi > 1/m (i.e., 1/5 = 0.20). MANOVA was performed on statistical package SPSS (version 11.0, SPSS, Chicago, IL, USA). Data are presented as means ± SE throughout the text and figures.
2. Results and discussion
Water temperature was 14.0–14.5ºC in the sampling brook. The snout–vent length of the collected 22 larvae was 22.0 ± 0.8 mm, corresponding to 0.10–1.10 (0.40 ± 0.06) g in wet weight according to Takahara et al. (2008).
Salamander larvae tended to inhabit two distinct environ- ments with regard to current velocity (Fig. 4). The col- lected larvae were classified into fast (6 individuals) and slow (16 individuals) current groups. The current differed significantly between two groups (Fast: 43.6 ± 2.3 cm/s;
Slow: 8.0 ± 1.2 cm/s; t-test, t8 = -13.64, P < 0.001), but the snout–vent length did not differ (t8 = -0.69, P = 0.51).
The surface area of the stone where a larva was collected differed significantly between two current conditions
Fig. 4 Relationships between current velocity and snout–vent length of O. japonicus larvae.
Fig. 5 Surface area of stone (a) and water depth (b) of two current microhabitat conditions of O.
japonicus larvae. Means ± SE are from 6 and 16 individuals for fast and slow current conditions, respectively. Asterisks denote significant differ- ence between two current conditions (t-test, P <
0.05).
Snout–vent length (mm)
Current velocity (cm/s)
0 10 20 30 40 50 60
10 20 30 40
(b) (a)
Water depth (cm)Stone area (cm2) 0 200 400 600
0 2 4 6 8 10
Fast Slow
*
*
Current condition
(Fast: 513.7 ± 50.5 cm2, Slow: 137.7 ± 16.6 cm2; t-test, t5
= -10.02, P < 0.001; Fig. 5a). The water column depth at sites where a larva was present also differed significantly between two current conditions (Fast: 7.5 ± 0.9 cm, Slow:
5.4 ± 0.8 cm; t8 = 8.31, P < 0.001; Fig. 5b).
Field survey showed that the compositions of ben- thic animals did not differ between fast and slow current conditions (MANOVA, Wilks’ lambda = 0.50, F4,1 = 0.25, P = 0.89). Mayfly nymphs and caddisfly larvae accounted for 82% (fast) and 70% (slow) of the total benthos com- munities (Fig. 6a, b). Terrestrial invertebrates (e.g., crane fly; categorized as “Others”) were collected only in the slow current condition. The composition of stomach con- tents did not differ between two current conditions (Wilks’
lambda = 0.88, F5,16 = 0.42, P = 0.83). The stomach contents were similar to the benthos communities in the brook, in that both mayfly nymphs and caddisfly larvae accounted for 88% (fast) and 70% (slow) in the total stomach contents (Fig. 6a, b). In addition, terrestrial invertebrates (moth larvae, flies, and mites; categorized as “Others”) were found from stomachs only in the slow current condition. This is consistent with the results of field survey in that terrestrial invertebrates were found only in the slow current condition.
The Manly’s Index showed that O. japonicus larvae selectively foraged on caddisfly and fly larvae in the two current conditions (Fig. 7a, b). Despite the high abun- dance of mayfly nymphs in the field (Fig. 6), feeding preference was not detected in the stomach contents (Fig.
7a, b). Salamander larvae may prefer prey animals with less mobility in natural environments, because caddisfly and fly larvae are less-mobile compared with mayfly nymphs. Salamander larvae selectively foraged prey items of the group “Others” in the slow current condition (Fig. 7b). This preference is due to terrestrial prey items found only in the slow current condition. Onychodactylus japonicus may catch occasionally flowing terrestrial insects because of slow current.
Listed in table 1 are the number of salamander larvae that consumed prey items in each category (N), and the proportion of salamander larvae that consumed prey items in each category (R). We also showed the prey animals that were identified to the family level (Table 1).
The caddisfly and fly larvae were frequently found from the stomachs (R = 0.68 for caddisfly and R = 0.59 for fly).
Our study on stomach contents suggests that natural benthos communities are of primary importance for the survival of O. japonicus larvae. Knowledge of the food habits in the natural environments is necessary for understanding life history of amphibian species (Ihara, 1998; Kohmatsu, 2001; Matsui et al., 2003). It has been reported that the frog Rana nigromaculata feed on larger prey animals as it grows (Hirai, 2002). In addition, prey
Fig. 6 Proportions of each prey category collected from the field and found in the stomach con- tents of O. japonicus larvae. Means of stomach contents were calculated from 6 and 16 individ- uals for fast (a) and slow (b) current conditions, respectively.
Fig. 7 Manly’s preference index (α) for the prey cat- egories of O. japonicus larvae in fast (a) and slow (b) current conditions, respectively. The α values above the broken line at 0.2 indicate that the prey is preferred.
ProportionProportion
0 0.1 0.2 0.3 0.4 0.5 0.6 0.70 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Stomach
(b) Slow (a) Fast
Mayfly
nymphs Caddisfly
larvae Stonefly nymphs Fly
larvae Others Field
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5
(b) Slow
Manly Index (α)Manly Index (α)
(a) Fast
Mayfly
nymphs Caddisfly
larvae Stonefly
nymphs Fly
larvae Others
items for the frog Hyla japonica seem to vary seasonally the number and species in natural environments (Hirai and Matsui, 2000). Therefore, such ontogenetic and sea- sonal changes in the diet of O. japonicus larvae need fur- ther investigation for their conservation.
Acknowledgements
We are grateful to Prof. M. Yuma of Ryukoku University and Prof. Z. Kawabata of Research Institute for Humanity and Nature for the opportunity to under- take this work. We thank Associate Prof. K. Nozaki of Sugiyama Jogakuen University for his comments on the manuscript. We also thank two anonymous reviewers for their critical comments and revision of the manuscript.
References
Blackburn, D. C. and Moreau, C. S. 2006. Ontogenetic diet change in the arthroleptid frog Schoutedenella xenodactyloides. Journal of Herpetology 40: 388–
Chesson, J. 1978. Measuring preference in selective pre-394.
dation. Ecology 59: 211–215.
Genkai-Kato, M., Mitsuhashi, H., Kohmatsu, Y., Miyasaka, H., Nozaki, K., and Nakanishi, M.
2005. A seasonal change in the distribution of a stream-dwelling stonefly nymph reflects oxygen supply and water flow. Ecological Research 20:
223–226.
Genkai-Kato, M., Nozaki, K., Mitsuhashi, H., Kohmatsu, Y., Miyasaka, H., and Nakanishi, M. 2000.
Push-up response of stonefly larvae in low-oxygen
conditions. Ecological Research 15: 175–179.
Goris, R. C. and Maeda, N. 2004. Onychodactylus japonicus (Houttuyn, 1782). p.36–38. In: Guide to the amphibians and reptiles of Japan. Krieger pub- lishing company, Florida.
Hansen, R. A., Hart, D. D., and Merz, R. A. 1991. Flow mediates predator-prey interactions between tri- clad flatworms and larval black flies. Oikos 60:
187–196.
Hayase, N. and Yamane, S. 1982. Life history during the aquatic life period of a salamander, Onychodactylus japonicus (Houttuyn), at Mts.
Tsukuba, Ibaraki, Japan. Japanese Journal of Ecology 32: 395–403 (in Japanese with English abstract).
Hirai, T. 2002. Ontogenetic change in the diet of the pond frog, Rana nigromaculata. Ecological Research 17: 639–644.
Hirai, T. and Matsui, M. 2000. Feeding habits of the Japanese tree frog, Hyla japonica, in the reproduc- tive season. Zoological Science 17: 977–982.
Hynes, H. B. N. 1970. The ecology of running waters.
Liverpool University Press, Liverpool, UK.
Ihara, S. 1998. The food habitats of Hynobius tokyoensis in broad-leaved forest floor: a research using stomach flushing method. Edaphologia 60: 1–9 (in Japanese with English abstract).
Kohmatsu, Y. 2001. Effects of head shape variation on diet selection in larvae of the salamander Hynobius retardatus. Current Herpetology 20:
63–67.
Kuhara, N., Nakano, S., and Miyasaka, H. 2000.
Flow rate mediates the competitive influence Table 1 Summary of stomach contents of 22 salamander O. japonicus larvae. N, number of salamander larvae that consumed prey items in the category; R, proportion of salamander larvae that consumed prey items in the category, i.e., R = N/22.
Prey category
Order N R Animals that were identified to the family level Mayfly nymphs
Ephemeroptera 11 0.50 Heptageniidae (7), Baetidae (2), Ephemerellidae (2), unidentified (4) Caddisfly larvae
Trichoptera 15 0.68 Rhyacophilidae (5), Glossosomatidae (2), unidentified (17) Stonefly nymphs
Plecoptera 1 0.05 Leuctridae (1) Fly larvae
Diptera 13 0.59 Chironomidae (23) Others
8 0.36 HydrophilidaeA (3), ChironomidaeA (1), CarabidaeA (1), AthericidaeL (1), TipulidaeA (1) LepidopteraL* (1), Acari U* (1)
Number of prey animals that were identified to the family level is shown in parentheses.
Asterisks (*) indicate prey items that were identified only to the order level in the category “Others”.
“A”, “L” and “U” indicate adults, larvae (or nymphs), and unknown, respectively.
of a grazing caddisfly on mayflies. Ecological Research 15: 145–152.
Matsui, K., Mochida, K., and Nakamura, M. 2003. Food habit of the juvenile of the Japanese newt Cynops pyrrhogaster. Zoological Science 20: 855–859.
Nakano, S., Kawaguchi, Y., Taniguchi, Y., Miyasaka, H., Shibata, Y., Urabe, H., and Kuhara, N. 1999.
Selective foraging on terrestrial invertebrates by rainbow trout in a forested headwater stream in northern Japan. Ecological Research 14: 351–360.
Nowell, A. R. M. and Jumars, P. A. 1984. Flow envi- ronments of aquatic benthos. Annual Review of Ecology and Systematics 15: 303–328.
Ogihara, M. and Nakamura, T. 1974. Distribution of Onychodactylus japonicus and Hynobius naevius kimurae in Yamanashi Prefecture. Proceedings of the Japanese Society of Systematic Zoology 10:
74–78 (in Japanese with English abstract).
Ohba, S. and Nakasuji, F. 2006. Dietary items of preda- cious aquatic bugs (Nepoidea: Heteroptera) in Japanese wetlands. Limnology 7: 41-43.
Poff, N. L. and Ward, J. V. 1995. Herbivory under dif- ferent flow regimes: a field experiment and test of a model with a benthic stream insect. Oikos 71:
179–188.
Stuart, S. N., Chanson, J. S., Cox, N. A., Young, B. E., Rodrigues, A. S. L., Fischman, D. L., and Waller, R. W. 2004. Status and trends of amphibian declines and extinctions worldwide. Science 306:
1783–1786.
Sugiyama, K. 2004. Onychodactylus japonicus (Houttuyn, 1782). p.192. In: Editorial Committee of Red Data Book of Shizuoka Prefecture (ed.), Red Data Book of Shizuoka Prefecture. Hagoromo shuppan, Shizuoka (in Japanese).
Takahara, T., Miyasaka, H., Genkai-Kato, M., and Kohmatsu, Y. 2008. Length–weight relation- ships in six amphibian species of Japan. Current Herpetology 27: 43–45.
Wheeler, C. A., Karraker, N. E., Welsh, H. H. J., and Ollivier, L. M. 2007. Diet of the Del Norte Salamander (Plethodon elongatus): differences by age, gender, and season. Northwestern Naturalist 88: 85–94.
木曽福島の黒川支流に生息するハコネサンショウウオ 幼生
Onychodactylus japonicus
の食性に関する予備的調 査髙原輝彦１，２）＊・加藤元海１，３）・宮坂 仁１）・神松幸弘１，２）
要 旨
2001年７月, 長野県木曽福島における黒川支流アカシオ沢に生息するハコネサンショウウオ Onychodactylus japonicusの幼生を採集し, 食性と生息場所の物理環境について調べた. ハコネサンショ ウウオ幼生の生息場所は, 流速の速いところと遅いところの2つに分かれたが, 流速によって胃内容物 とアカシオ沢における餌動物群集の組成比に違いはみられなかった. ハコネサンショウウオ幼生は流 速の速いところと遅いところの両方でトビケラとハエ目の幼虫を選好し, 流速の遅いところでのみ陸 生昆虫を含む餌動物のグループを選好していた. ハコネサンショウウオ幼生は, 底生動物を主要な餌資 源として利用しており, 流速が餌の選好性に影響を及ぼしていることが示唆された.
キーワード：餌選好性, 底生動物, マンリーの選択性指数, 物理環境, 陸生昆虫